A System with Intelligent Editing for Extracting Ridge and Ravine Terrain Features
Greg S. Schmidt, J. Edward Swan II, Derek Overby, Lawrence Rosenblum, and Erik B. Tomlin. A System with Intelligent Editing for Extracting Ridge and Ravine Terrain Features. Memorandum Report Naval Research Laboratory, 2003.
Download
Abstract
We describe a system for extracting ridges and ravines from elevation data. The application context is a map-based military planning tool which allows users to select ridges and ravines by simple mouse clicks. The extracted terrain features are complete in the application-specific sense that they conform to what our users expect a single ridge or ravine to look like. Supervision is supported by a graphical user interface, which allows an analyst to modify algorithm parameters as well as perform intelligent mouse-based editing operations. Among similar existing systems, ours is unique in that it focuses on the three classically difficult operations of (1) combining partial features, (2) splitting multiply-connected features, and (3) removing micro-terrain features for extracting high-level structure from classified pixels.
BibTeX
@TechReport{TR03-rrtf,
author = {Greg S. Schmidt and J. Edward {Swan~II} and Derek Overby and
Lawrence Rosenblum and Erik B. Tomlin},
title = {A System with Intelligent Editing for Extracting Ridge and
Ravine Terrain Features},
institution = {Naval Research Laboratory},
type = {Memorandum Report},
year = 2003,
abstract = {
We describe a system for extracting ridges and ravines from elevation data. The
application context is a map-based military planning tool which allows users to
select ridges and ravines by simple mouse clicks. The extracted terrain
features are complete in the application-specific sense that they conform to
what our users expect a single ridge or ravine to look like. Supervision is
supported by a graphical user interface, which allows an analyst to modify
algorithm parameters as well as perform intelligent mouse-based editing
operations. Among similar existing systems, ours is unique in that it focuses
on the three classically difficult operations of (1) combining partial features,
(2) splitting multiply-connected features, and (3) removing micro-terrain
features for extracting high-level structure from classified pixels.
},
}